Power, Uncertainty, and War Duration

This research note examines a key implication of informational theories of crisis bargaining—that asymmetric adversaries generate more uncertainty than symmetric ones—through an empirical analysis of 188 National Intelligence Estimates. Using sentiment analysis and keyword assisted topic modeling, I find that intelligence analysts perceive symmetric adversaries to be at least as uncertain as asymmetric ones, even as analysts perceive the former as more threatening. The psychology of uncertainty assessment offers insight into the gap between theoretical and perceived uncertainty. I connect these observations to a puzzle in global patterns of conflict duration, namely why asymmetric wars have grown longer over time, while symmetric ones have not.